CA1306349C - Polyester fiberfill - Google Patents

Abstract

ABSTRACT
IMPROVEMENTS IN POLYESTER FIBERFILL
Polyester fiberfill having spiral crimp that is randomly-arranged and entangled in the form of fiberballs with binder fibers, preferably with a minimum of hairs extending from the surface of the fiberballs, so as to be air-transportable on account of the low cohesion between the balls. A process for making such fiberballs by repeatedly air-tumbling small tufts of such fiberfill/binder/blend against the wall of vessel.
Improved bonded batts or molded articles or other bonded articles obtained by bonding such fiberballs.

Description

~3~)163~

TITLE
IMPROV~MEN~S lN POLYEST~R F~B~RFILL
TECHNICAL FI~LD
Thi~ invention concerns i~provement~ in and relating to polyester fiberfilling material, ~ommonly referred to as polyester fiberfill, and more particularly ~o providing polyester fiberfill in the ~orm of ~iberball~
~ontaining binder fiber6, that may be bonded to provide u6ef~1 new through-bonded products, and to proces6es for preparing these new products.
8~C~GROUND OF INVENTION
~ her~ally-bonded (polyester) fiberfill batts (or battings) ~re well known ~nd have gained large scale co~mercial use, particularly in Europe. sinder fiber~ can be blended intimately into ~he fiber~ill to achieve true ~through-bonding" of fiberfill batts, and thu~ achieve better dur~bility versus resin-bonding, which was the conventional method, and can also provide reduced flammability versus resin-bonding. Such binder fiber blends are used on a large 6cale in furnishings, mattresses ~nd 6imilar end-uses where stronq support is desired. ~owever, they are seldom used as the only illing ~aterial in these end-uses, particularly in ~urnishing ~e~t cu6hions, where the common practice is to use the fiberfill batt~ as a "rapping~ Por a fo~ core.
It is believed that the main reason is probably that, to obtain the desired resilience and performance in 100~
fiberfill cushions, it would be necessary to provide such relatively high density ~s has hitherto been considered too c06tly ~nd difficult with the present techniques, and as misht not provide desirable performance aesthetically.
In a conventional fiberfill batt, the fibers are arranged in parallel l~yer~ whizh are bonded together. In 6uch a layered structure, ~ny pressure ~pplied during u6e ~s a cushion is essentially perpendicular to the direction of the fiber6 and I believe that may be at least partly why 3~ ~

. .

: ` ~
.

-2- ~3~3~3 6uch ~ high density must be reached to achieve the de~ired resilience and durability using conventional lnyering and bonding technique~.
SUMMARY OF INVENTION
According to the invention, there are provided new fiberfill 6tructure6 that may be bonded to prov~de product~ of improved performance, c6pecially with regard to resilience ~nd durability, over what has been ~vailable oo~mercially hitherto, ~6 will be explained hereinafter.
According to one aspect of the invention, there are provided fiberballs of average dimension ~bout 2 to about 15 mm, consisting essentially of randomly~arranged, entangled, ~pir~lly-crimped polye~ter fibesfill having a cut l~ngth of ~bout 10 to about 100 ~m, intimately blended with binder fiber6 in ~mount about 5 to about 50~ by weight o the blend. Alternatively, there ~re provlded fiberball~ of average di~ension about 2 to 15 ~m, consi~ting e6sentially of randomly-~rranged, ent~ngled, 6pir~11y-crimped bicomponent polyeQter/binder ~ateri~l fibers, having a cut length of ~bout 10 to about 100 mm.
According to another a~pect of the invention, there i6 provided 3 process for making polyester fiberballc from an intimate blend of spirally-cri~ped polye~ter fiberfill and of binder fiber~, wher~in small tuft~ o~ the blend are repeatedly tumbled by air ~gain~t the wall of a ve~6el to provide the fiberball6.
Alternatively, there is provided a process ~or making polye~ter fiberball~ from spirally-crimped bicomponent polyester/binder material fibers, wherein small tufts of the ~pirally-crimped fiber6 are repeatedly tumbled by ~ir against the w~ll of a vessel to pro~ide the fiberb~
-- Aocording to further ~spect~ of the lnvention, there are provided entirely new resilient shaped articles or structures consi~ting es~entially of thermally-bonded, ~pirally-crimped polyester fiberfill, and proce66e~ for making these b~nded products from the fiberball~ bf the ~3~63~
lnvention. These a6pect6 will be dealt with in ~reater detail below. A6 will be ~een, the f~berb~lls of the invention open entirely new po~ibilitie6 and the u6e of alternat$ve t~chniques for preparing bonded ~rt~clo6 rom polyester flberfill, which, hitherto, ha6 been ll~lted, effectively, ln co~mercl~l practice, to the u~e of carded web~ ~nd batt~, and bonding and ~haping ln the for~ of a batt, with all the con6traint6 that thi~ has impo~ed in pr~ctice.
~RIEF DESCRIP~ION OF DRA~INGS
Figures l and 2 are enlarged photograph~ of ~iberball~ ~ccording to U.S. Patent No. q,618,531.
Figure~ 3 and 4 are ~che~atic dr~wings in ~ection of the ~chine ufied to ~ake the ~iberball~ in the Example6 herein.
DETAILED DESCRIP~ION OF T~E INVENTION
So~e idea of the nature of the fiberball~ of the invention, ~nd e6pecially of the nature of the configuration~ t~ken up by the ~pirally-crimped fiberfill therein, can be gained from Figures 1 ~nd 2 of the accompanying dr~wings. ~or convenience, at th~ point, reference i~ made to ~y copending application, i.e. U.S.
P~tent No. 4,618,531, directed to réfluff~ble flberballs of fipirally-crimped polyester fiberfill, and to ~ Droce for ~aking ~uch fiberball6. The objective of my copending ~pplic~tion wah to provide a ~ynthetic product as a real alternative to down, in the ~ense o~ h~vlng refluffable characteri6tic~ ~vailable from down) 3nd al~o with wa~h~bility (unl~ke down) and At ~ lower co~t than down. A~ indicated, thi6 objective wa~ obtained by providinq refluffable fiberb~ from ~p;r~lly-crimped polye6ter fiberfill. An e66ential element wa~ the u6e of 6uch 6pirDlly-cri~ped fiberill. Such refluf~ble ~iberball6 can be obtained by ~ir-tu~bling s~all tufts of fiberfill (having ~pirDl crimp) repe~tedly ~g~in~t the ~ .. . . . .

~4~ ~3~63~

wall of a ve6sel as illustrated in Flgure~ 5 ~nd 6 of my copending applicati~n, corr~sponding to rigures 3 and 4 herein. The objective of the pre~ent invention i~
entirely different from the objective of my cop~nding ~pplication, a~ indicated above. Moreover, the 1berballs of the pre6ent invention are distingui~hed ~rom the refluffable ~iberball~ ~pecifically di6c1062d in ~y copending application, by the content of binder fibers, to achieve the bonding and the new bonded products that are the objective of the present invention. Neverthele6s, the technique6 used for ~aking fiberballfi are similar, and es6entially the 6ame apparatu~ ~ay be u6ed in both in6tanc~6, and Figures 1 ~nd 2 ~ay be h~lpful $n visualizing the fiberball6 of the invention, and the spirally-crimped fiberfill therein.
A~ indicated, an essential element of the present invention is the use o~ ~iber6 hDving ~ignificant curline~6, 6uch as is referred tG herein a~
~pirally-crimped fiberfill. Such fibPrs have ~ rmemory"
that provides them with a natural tendency to curl, i.e.
to take up helical or ~piral configurations. ~he provi6ion of ~uch spiral criMp i6 itself well-known for other purpo6es. Thi~ can be provided economically by asymmetric-jet-quenching of freshly-extruded poly2ster filaments, a~ t~ught, e.g. ~n ~ n U.S. ~at. No~.

3,050,821 or 3,118,012, especially for filament6 of drawn denier in the range ~bout 1 to 10. The ~pir~l crimp i6 believed to re6ult from difference6 in crystalline ~tructure acros~ the cross-~ection of the fiber~, which provide differential ~hrinkage, 60 the fiberfi c~rl helically upon appropriate heat-treatment. ~uch curls need not be regular, and in fact ~re often quite irregular, but are generally in 3 dimen~ion6 and ~o ~re re~erred to as 6piral crimp to di~tinguifih from the es~entially 2-dimen6ional saw-tooth crimp induced by ~echanical ~eans, ~uch as a ~tuffer box, which i~ the 3063~

preferred msthod u~ed commercially for crimping polye6ter tow precursors to ~taple fiber at thi6 tims.
Asymmetric-jet quenching is the ~echnique that was u~ed to ~ake the fiberballs in E~amples 1 5 herein. An alternative way to provide ~piral-crimp i~ to ~ake bicomponent filaments, sometimes referred to ~ conjugate filaments, whereby ~he components have different shrinkages upon being heat-treated, and so become spirally-cri~ped. sicomponent6 are generally more expen~ive, but may be preferred for ~ome end-uses, e~peci~lly if it i~ desired to use fiberfill of relatively high denier, ~uch as i6 more difficult to ~piral-crimp adequately by an asymmetric-jet-quenching technique.
Bicomponent polye~ter il~ments are taught, e.g., $n ~vans et al. U.S. Pat. No. 3,671,379. ~articularly gosd results have been achieved by using a bicomponent polyester fiberfill ~old by Unitika Ltd. as H38X, referred to in ExamRle IIIs of copending application ~P Al 0 203 469. Of course, especially with bicomponent filament6, there i6 no need to use only polyester components. A ~uitable polyamide/polyester bicomponent filament can be 6elected to give a good ~piral-crimp. Still further method~ of obtaining fiberfill with a "memory" and ability to crimp spirally are disclosed in Nippon E~ter Jap~nese Patent Application ~okai No. 57~56512, publi6hed ~pril 5, l9B2, and in Toyo ~oseki U.~. Patent No. 1,137,028, which indicate that hollow fiberfill can be obtained with this property.
Apart from the ~piral-crimp, which i6 essential, the fiberfill staple fiber6 ~ay be solid or hollow, of round cro6s-section or non-round, and otherwi~e ~s disclosed in the prior art, according to the aesthetics desired and 3ccording to what materials ~re available.
The spiral-crimp ~ust be developed in the ~iberfill ~o that making the fiberball6 beco~es po~sible.
Thus a tow of asymmetrically-jet-quenched polye~ter -6~ ~3~3~

filameMts i~ prepared by ~elt 6pinning ~nd gathering the ~pun filament~ together. The tow ifi ~hen dr~wn, optionally coated with a 6urface modifier, optionally relaxed before cutting conventionally to form ~t~ple ~ibers, ~nd preferably relaxed after cutting to enh~nce the asym~etric chnracter of the fibers. Thi6 ch~r~cter is required ~o the fiber~ will curl ond form the de~ired fiberballs with minimal hairine6~. Convention~l mechanical crimping, ~uch a8 by a ~tuffer-box technique, is not generally desired because inappropri~te heat-treatment can de~troy the de6ired fipiral-crimp, and ~o cuch mechanically-cri~ped fiberfill would not form fiberball~, as de~ired. Such mechanical crimping i6 not an ~lternative to spiral-crimp, becau6e mechanical crimping gives a ~aw-tooth crimp which will no~ form the desired fiberballs. ~owever, we have found that fiberballs c~n be obtained if so~e 6uitable degree of ~echanical erimp with appropriate heat txeatment i~
provided to the precursor filamentary ~ow, in whieh case the eventual fiberfill will have a configuration that is a result of combining thi~ ~echanical crimp and 6piral crimp. This is the techniq~e u6ed in Example6 6-10 herein. We r~fer to this crimp a6 ~-cri~p (omega-crimp) because the configuration of the fiber~ rese~ble6 the shape o~ this Greek letter ~, being a combination of a ~aw-tooth from the mechanioal cri~ping 6uperimpo6ed on the curl of the 6pi ral crimp obtained becau6e of the "~emory"
referred to above. ~his ~-crimp ~ay be obt~ined in other w~y~ .
An e~6ential ele~ent of the fiberball~ o the pre~ent invention i~ the binder fiber~, which are preferably used in amount about 5 to ~bout 50~ by weight of the blend, the preci6e amount depending on the ~pecific constituent6 ~nd the desired end-u~e, but about 10 to about 30% generally being preferred. A6 indic~ted above, binder fibers are well known and have been u6ed -7~
~30~

comm~rcially f~r obtaining ther~ally-bonded batts of polyester fiberfill. Such conventional binder fibers, e.g. of lower ~elting polye~ter, ~y be u~ed ~ccordiny to the present invention afi 6uch, or ~odified ~ppropriately.
Several option~ ~re, however, a~ail~bl~ w~ll be clear bereinafter. The gener~l requirement~ for blnder ~ibers are conveniently ~et out ln ~am~ V.S. Pat~nt No. ~,281,042 ~nd Franko~kv U.S. Patent No. 4 3~4 ~17- As indicated therein, and di~cu~sed herein~fter, dependin~ on the intended end u6e, it may be preferred ~o provide blends of binder fiber w~th surface-modified ~61ickened) fiberfill (to provide ae~thetics that ~ay be de~ired in the ther~ally-bonded product), including triple blends ~lso with unslickened fiberfill (-f de~ired to provide bonding sites, when the ~lickened fiberfill i~ not ~o ~menable for thi~ purpose~ ~c well ~z the binder f$ber~
themselves. An i~portant require~ent of the binder material i6 that it h~ve a bonding temperature lower than the ~oftening temperature of the polyest~r fiberfill.
Thus the binder 6ho~1d be of appropriately lower ~elting poin~ than the polys~ter ~iber, e.g~ ~ome 20C or 30C, or prefer~bly 50~ lower, depending on the sen6itivity of the ~terials to heat ~nd th~ efficioncy of thé bonding equipment and condition6, ~o that ther~al bonding of the blend ~ay t~ke pl~ce conveniently without deleteriously ~ffectin~ the physical proper~ies of the po~y~6ter fiberfill itsel~, or be otherwise capable of being ~ensiti~ed ~o a~ to provide it~ essential function o~
bonding the polye6ter fiberfill. It will b2 under~tood that, if the binder fibers are monocomponent fiber~ in the blend, they may 106e their fiber for~ during the bonding operation, ~nd thereafter the binder ~ay exi~t ~erely as globs binding the inter6ections of the polye6ter iberill. If, however, the binder fiber6 are bicomponent fibers, e.g. if preferred sheath-core fibers are u~ed, and - -~L3~3~
only the 6heath comprising e.g. about 5 to about 50% of the bicomponent i~ ~ binder materi~l, wherea~ the core i~
a higher ~elting component that can remain in fiber ~orm after the bondin~ operation, then the final bonded product will comprise the6e remaining core element~ from the original binder fiber6 in addition to the polye6ter fiberfill. Indeed, it ~ay be pos~ible and desirable to provide a ~ulticomponent binder fiber that i~ al~o spirally crimped ~nd so can by it~elf perform all the requirement~ of the present invention. In other words, there would be no need for ~ blend of ~epara~e binder fibers and 6pir~11y-crimped fiber~, but the fiberball~ of the invention would consi~t es~entially of ~pirally-cri~ped, multico~ponent, binder fiber~ that are fir~t formed into the fiberball6, and then at a later stage treated ~o to activate the binder material component, thereby leaving a bonded assembly or ~haped article of bonded fiberfill.
The binder fibers are preferably of ~i~ilar dimensions and processing characteristics to the polyester fiberfill, to permit easy intimate blending, although this i~ not e~enti~l, and may not even be desirable depending on the intended final u6e and the components. For instance, if the binder fiber i~ a bicomponent, u~ed in rel~tively large quantitie~ may be desirable that the final bonded product comprifie bonded fiber~ of e~senti~lly similar dimensions and characteristic6. A5 indicated, it may be advantageous to provide the binder fiber in spirally-crimped form. ~hi6 will be partlcularly desir~ble if the binder fiber compri~es a ~igni~icant or large proportion of any blend, so ~6 to facilitate the formation of the fiberball~, although it is possible for 6pirally-crimped fiberfill to for~ ~atisfactory fiberball6 even in the pre6ence of other fiber~ that are not spir~lly-crimped, and 80 dilute the effect of the 6pirally-crimped component~.

~L3q:~63'a~
g Bearing the above in mind, the selection of the various characteristics, amounts and dimensions of the fiber constituents will depend generally on the intended end use, and the aesthetics of the bonded article, and such considerations as cost and availability. Generally, the dtex will be between 1 and 30, preferably at least 3 dtex, and preferably less than 20 dtex, and often approximately 5 dtex or up to 10 dte~, and the cut length is generally about 10 to about 100 mm, preferably at least 20 mm and preferably up to 60 mm.
As indicated, it may be desirable to slicken (lubricate the surface) at least some of the fibers, and to use a conventional slickening agent for this purpose. This may be desirable for several reasons, e.g. for aesthetics in the final bonded product, and to improve durability, and also to reduce the cohesion of the fiberballs, and to permit them to be transported, e.g.
by blowing. If a conventional silicone slickener is used, however, this will reduce the ability of the ~lberfill to bond, and increase the flammability, as disclosed already and in copending C~nadian patent application No. 549,793 filed 1987 October 20, and so, preferably, the fiberfill will be coated with a hydrophillic slickener consisting essentially of chains of poly(alkylene oxide) as disclosed therein.
Several such materials are disclosed in the literature. Preferred materials are "curable" to the polyester fiberfill. ~or instance, a segmented copolymer of poly(ethylene terephthalate) and poly(ethylene oxide). Some such materials are available commercially, such as the textile ~mishing agent sold under the trademark ATLAS* G-7264 by ICI Specialty Chemicals, Brussels, but it may be preferred to use materials with less fiber to metal friction, as well as a low ~lber to fiber friction. Another material is sold as ZELCON*4780, by E.I.
du Pont de Nemours and Company. Other materials are * denotes trademark . .

-lo- ~3063~

dicclosed ~n Raynolds U.S. Patent No. 3,981,B07. ~veral seg~ent~d eopolye~tor6 con~i~ting e6~entially of poly(~thylene terephthDlate) ~egments ~nd o poly(~lkylene oxide) 6eg~ent~, derived from a poly(oxyalkyl~ne~ haYing ~olecular ~eight of 300 to 6,000 ~nd di~per6ions thereof are di6closed in McIntyre et ~1. U.S. P~tant No~. 3,416,952, 3,557,039 and 3,619,269, and in vasiou~
other patent ~peeificat~on~ di6closing like cegmentod copolymers oontain~ng poly~ethylene terephthalate) ~egment~ and poly(alkylene oxide) ~egment~. Generally the poly(alkylene oxide~ will be 4 poly(ethylene oxide), which i~ a ~atter of com~ercial convenience. Other 6uitable ~aterial6 ~nclude modified poly(ethylene oxide)/poly(propylene oxide) qrafted with functional groups to permit cro6slinking, e.g. by treatment wi~h citric acld, ~uch d~ are available commercially fro~ Union C~rbide a6 '~CON~ 3207A. Other ma~eri~l~ th~t may include particularly ueeful co~po~ition~ ~re di6clo6ed ln Tei~in EP 159 882 and in ICI Amer~e~s EP 66944. Choice of a particular ~lickener will depend on the de~ired end-u~e, and many of the $ndicsted ~lickener~ differ in thelr ability to lubricate, e.g. to lower fiber-to-fiber and/or fiber-to-~et~l friction and a~ount~ of anion group6. If, for example, ~oi~ture tran~port and durability are deEired, but ~oftnes~ i~ not ~o important, item 12 ~n EP
6S944 ~ay be defiir~ble. Depending on the ~e~thetlc6 de6ired, the amount of ~lio~ener ~ay be ~djusted, between about 0.05 and about 1~, preferably about 0.15 to &bout 0.5%, on th~ weight of the fiberfill, being generally de~irable, depend$ng on, e.g., the ~ype of ~lickener and the effect desired.
Polye~ter fiberfill, like other 6taple fiber, has been generally transported in co~pres~ed bales, which ~re convention~lly fir~t treated in ~n opener, ~o a~ to ~eparate the individual f iber6 to so~e extent before they ~re further proce~sed, e.g. on a card if a parallelized * denotes trade mark ~3~63~
web is desired. For making products o~ the invention, it is not necessary, ~nd is generally undesirable, to completely parallelize the fibers, but it i6 desirable fir~t to open and 6eparate the fiber~ into di6crete tuft6 before treatment to form the fiberball6, æ~ will be described.
The fiberballs are formed by air-~umblinq 6mall tufts of fiberfill (having spiral crimp) repeatedly against the wall of a vessel ~o a6 to densify the bodie~
nnd make them rounder. The longer the treatment, generally the denser the resultinq balls. It i6 believed that the repeated impact~ of the bodies cause the individual ibers to entangle more and lock together because of the curl of the ~piral crimp. In order to provide an easily-transportable product, however, it i5 ~160 preerred to reduce the hairiness of the b~
because the spiral-crimp of any protrudlng fiber6 will rai~e the cohesion between neighboring fiberball6. ~his cohesion can also be reduced somewhat, however, by thorough distribution of a slickener, as described herein, to increase lubricity between the fiberballs. The 61ickener also affects the aesthetics. Depending on the aesthetics desired, the amount of tumbling and applacation of ~lickener may be adjusted.
The fiberball~ of the prefient invention comprise fiberc that are r~ndomly-arranged, as ~hown ln Figures 1 and 2, 6howing desirable light fluffy ball~ with low cohesion, because of the use of spirally-crimpet fiberfill. In contrast, a mass consi~ting only of regular polyester fiberfill, ;.e. ~echanically crimped polyester fiberfill without any 6pirally-crimped material, cannot be formed into balls by the process of the invention. Such regular fiberfill, like other fibers, ~uch afi wool, can be forced into den6e ~s~emblies, includinq balls, by u~ing very high shearing forces. ~hese dense assemblies ~re entirely different from the fluffy blowable iberballs of ~6~
the present invention, being harder, denser ~nd hairy and ~re ~t desirable for the purposes of the present invention.
The air-tumbling has been 6atisfactorily performed in ~ modified machine, based on ~ Lorch ~achine as described in ~y copending application, now U.S. Patent No. 4,618,531, and ~s illustrated ln Figure6 3 and 4 herein. This machine was used in the ~xample herein.
The resulting fiberballs are easily transported, fo! instance, by blowing, e~pecially if the hairiness is reduced by increasing lubricity, aS described herein and in my copending ~pplication.
The6e fiberball~ may then be compressed and bonded together to form bonded 6tructure~ that ~ay ~uperficially re6emble bonded batts or molded into any desirable ~hape. For instance, the iberball6 ~ay he blown into a light tickin~, or a non-woven, ~nd then heated to produce a cushion-like article in the chape of the ticking. As a result, the final product has improved resilience and performance, ~s indicated hereinafter, and is very different from prior art bonded batts. It is believed that the improvement results from the fact that the fibers have a ~igni~icant component in every direction, as contrasted with the primarily parallelized fibers of pr~or art layered batts. The difference ln performance is 6urpri~ing and siqnificant, as can be ~hown by examining the different 6tructure~ when they are ~upporting a weight. The bonded products of the invention act like many independent ~prings that support the weight above them, wherea~ the parallelized fibrous structures of the prior art will pull inwards from the ~ides, for reasons that can be rationalized in retrospect. Another advantage is the faster moisture transport, which i6 believed to result from porosity between the fiberballs, which i~ o particular potential interest for 6tructures 6uch a~ cushions and ~atresses wherein the principal or ~3~63~'13 only stuffing material i6 6uch fiberballs. The moisture transp~rt ~har~cteristios can be ~urther enhanced by the use of a permanent hydrophilic fini~h, ~s indic~ted.
Thus, the major expected end uses for the final ~tructures are for furni~hing cushions, car ~eat6, matre66e6 3nd like products, Such 6tructures may, if de6ired, be ~olded initially into the form finally desired by heating to activate the binder fiber in the fiberball6 within ~
ticking within a mold of the desired shape. Or the bonded structure may be formed in lon~ lenqths like prior art bonded batts, or in other 6tandard ~hapes, and then be cut and, if neces~ary, be reshaped as desired. Greater flexibility in this regard is available than with prior art bonded batt~.
Moreover, it may prove feasible to u~e the fiberballs of the invention in a manner completely different from that commercially used heretofore with prior fiberfill products, namely by bonding the fiberballs individually in a fluidized bed, ~nd then blowing the individual balls into a ticking. ~he resulting new product is refluffable, and 60 en~irely different from prior Drt bonded fiberfill products, but more like cushions filled with feathers and chopped foam.
Such new product has, in addition to good resilienee and durability, the novel char~cteri~tic that the individual balls can move in ths ticking in a similar manner ~o down and feather blends. In 6uch products, it is ~gain desirable to reduce cohesion by application of appropriate lubricants or clickeners for this purpose (and for promoting moisture tran~port, as disclosed in my copending application). Thi6 reduction of hairine6s/cohe~ion improves the transport~bility of the fiberball6, e.q. by blowing, and improves the 60ftness of the product in end uses where this i6 desirable, and also offers ~n lmproved degree of moisture tr~nsport that is believed unattainable with prior products. In 6uch product6, the di~ensions of ~3~)~3'~
the ~iberballs are believed i~port~nt for a~sthetlc reasons, ~6 describe~ ln U.S. Patent No. 4~618,531, av~rage dimensions of about 2 to about 15 mm being preferred.

DSSCRIPTION OF TEST METHODS
~ ulk ~easure~ents were ~ade conve~ionally on an Instron ~achine to measure the compre~6ion forc~G and the height o~ each 6~mple cushion, which w~s compre6sed with the appropri~te foot of diameter 10 cm ~ttached to the Instron. From the In~tron*plot are noted ~in cm) the Second InitiAl Height (IH2) of the test matexial, i.e. the height ot the beginning of ~he 6econd compression cycle, the Support Bulk (S~ 60N), i.e. the hei~ht undor a co~pression of 60N, ~nd the Bulk ~heiqht) under a compres6ion of 7.5N, ~ 7.5N). The ~oftne~s i~ calculated both in ab~olute terms (AS, i.e. I~12-~ 7.5N) Dnd in rela~ive terms (RS - a~ a percentage of IH2). ~he fir~ness of ~ cushion correlate6 with 6trong ~upport, i.e.
the Support Bulk, and inversely with softns6s.
Resilience i~ measured a~ Work Recovery (WR), i.e. the ratio of the ~rea under the whole recovery curve calcul~ted ~s a percentage of that under the whole compre~sion curve. The higher the ~R, the better the resilience.
Dur~bilitv - ~ach 6ample cushion was covered with ~ fabric having ~n ~ir per~eDbility of about 100 l/~q~./6ec and it6 compre~sion cur~e was ~essured and recorded ~ 8F (before flexing). The fir~er cu6hion6, whose test result~ ar~ ~hown ~n T~ble~ 2-5, were then ~ubmitted to 10,000 6uccessive flexings under ~ pressure of 13 kPa ~about 133 g/6q.cm.) at ~ rate of ldOO
cycles/hour ~nd the compression curve measured again and recorded a~ AF (after fl0xing) ~o ~s to ~how ~ny changes in bulk and resilience resulting from the flex tes~, as percentages (h). The pillows of Examples 15 et ~q. were flexed differently, as descrihed later, in relatlon to Table 6.

* denote~ trade mark - - -~1 S- ~3~6;3~

The invention is further described in the following Examples. All parts and percentages herein are by weight, and with respect to the total weight o iber, unless otherwi~e 6tated.
Example 1 A tow of asymmetrically-jet-quenched drawn poly(ethylene terephthalate) filaments of 4.7 dtex was prepared conventionally without mechanical crimping, using ~ draw ratio of 2.8X. The tow was cut to 36 ~m cut length and relaxed at a temperature of 175C to develop the 6piral cri~p. The 6taple was blended in the ratio of 80/20 with a 6heath/core binder fiber, cut to the ~ame cut length, ~nd having a 4.4 dtex. The blend was opened on a commercial opener and the resulting opened blend ~as proces6ed for 6 6econds on ~ Trutzschler cotton beater to separate the fibers into discrete small tufts. A batch of the resulting product was blown into the ~odified Lorch machine, as described and illustr~ted, and proce~sed for 1 minute at 250 rpm, then for 3 minutes at 400 rpm to convert the tufts into consolidated fiberballR.
The fiberball6 were packed to different extents, to provide a series of different densities fro~ 20 Rg/m3 (~) to 5D Xg/m3 ~E), as indicated hereinafter, into a box ~mold) ~de of wire me~h reinforced with 2 mm thick stainles6 steel bars with a rectangular base of 40 x 33 cm and where the height can be varied between 1 ~nd 25 cm.
Each sample of fiberballs wa~ compres6ed to n 6imilar height of about 9 cm, while varying the re6ulting density by changing the quantity of fiber ball~ packed into the box. ~he mold was then placed in an oven with an air flow ~cross the rectangular ba~e ~t a temper~ture of 160C for 15 minutes. After cooling the mold, the resulting "cushion" was released and the co~pression characteristics were determined, and are recorded in the top part of Table 1 as Items A-E. This indicates that the ~upport obtainable from products of the invention can be varied -16- 13~63~

over a wide range, by vary~ng the density, and ~h~t excellent sesilience (W~) is also obtained, especially at higher densitie6~ The durability is also excellent; (this i~ ~easured and di~cu6sed hereinafter in relation to h~gher density (firmer) products, with reference to Table 2). For comparison, 6imilar compression measuremen~s were made and are recorded in the lower part of Table 1 for 5 conventional ~aterials bonded under exactly the same procedure 86 in Example 1. The compositions of the~e ~Compari~ons" was a follows:
1. A triple 60/20/20 blend, compressed to ~bout 20 ~g/m~ (for compari60n with Item ~ of the invention), using the same binder fiber ~of dtex 4.4) in ~he same amount 20~, but containing 80% of commercial poly(ethylene terephthalate) fiber ill of three time~ higher denier (13 dtex), which would normally give much better resilience and more firmness (support bulk) than products from lower denier fibers), one guarter of which wa~
glickened with a commercial silicone slickener (20%), while the remaining three quarters ~60%) was "dryn, i.e.
unslickened.
2. An 85/15 blend, compressed to ~bout 25 Kg/m3 (for compari~on with Item B of the invention), of the same binder fiber ~15%) of denier 4.4 dtex, but cont~ining 85 of dry hollow commercial fiberfill of 6.1 dtex (~ignificantly higher than the 4.7 dtex fiber~ill u~ed in Item B).
Despite the lower dtex in the cuchions of the invention, Items A and especially ~ showed equal or better resilience (higher WR) and better support bulh (lower RS) than the comparisons 1 ~nd 2 of ~imilar density.
Furthermore, the products of the ~nvention have excellent durability,, whereas the compari~ons are much inferior especially in this respect. For higher densities, ~imilar compari60n blends would fare even worse, so I te~ted the following representative products used in furnishing seat ~3~63~

cu~hions or ~attress core6; (the characteristics of polyurethanes can be varied by changing the ingredients to increa~e the softness or firmness, so these qualities are not controlled merely by the density):-3. Commercial polyur~thane "soft" f~am core at35 Xg/m .

4. Commercial polyurethane "firm" foam core at 30 kg/m .

5. Commercial late~ core ~10 cm height) at 72 ~g/m .
The results in Table 1 indicate that products C
and D of the invention are comparable in resilience to the foam cushions 3 and 4, which are fir~er, ~nd product ~ o~
the invention is somewhat more resilient than the l~tex.
This is a significant achievement and could open the way for fiberfill ~o be used as the only filling material ~n certain end uses where previously foam cores have been used.
~ he durability of sample cushion E (at 50 ~g/m3 ) i6 recorded as Example 1, in Table 2, and is compared with cushions of ~imilar density made as described in Examples 2-10~
Examp~e 2 ~ he procedure of Example 1 was ~ollowed, except that the fiberballs were mixed with 10~ of the ~me binder fiber ~efore belng molded at 50 Rg/m~ to glve a product of somewhat higher resilience and lower bulk lo~ses, i.e.
some~hat better durability.
Example 3 ~ he procedure of Example 1 was followed, except th~t the fiberballs were treated with 0.35~ of 3207A UCON
and dried at 50DC before being molded. This product shows lower initial resilience but 1 ss loss of bulk or resilience after the durability test.
Example 4 The procedure of Example 3 was followed, except that 0.35% of G-7264 was used instead of 3207A UCON. This -lB- ~3063~3 product shows equal bulk and lower resilience than Example 1.
Example 5 The fiberballs of Example 4 were mixed with 10%
of the same binder iber in random form (not in balls) as in Exa~ple 2 before molding. Thi~ product shows the best co~bination ~f durability of resilience with good bulk.
Summarizing the durabil$ty results of Examples 1 to 5, Example 1 ~hows "dry~ fiberball~ molded alone, whereas Examples 3 and 4 show ~iberball~ slickened with non-silicone PEO-type ~lickeners molded alone, ~xample 2 shows dry fiberballs ~ixed with random binder fiber before molding, while Example 5 shows a eombin~tion of this feature and of the ~ore effçctive slickener of Example ~. A8 ~hown in Table 2, the slickened items of Ex~mple~ 8 ~nd ~ performed remarkably well, ~howing that good bonding occurred, ~nd held up well throughout the ~lexinq treatment, despite the coating with these particular slickeners (whereas silicone-slickened fibers do not bond). Indeed their durability was better at equal support bulk than dry Example 1, but the resilience was lower. The best refiults were in Example 5, whers the resilience was almost the same initially, but better ~fter the durability test, and the ~upport bulk showed better durability.
Examples 6-10 These Examples correspond to Examples 1-5, respectively, except that the tow of 4.7 dtex was mechanically crimped (to provide a ~ild mechanical crimp in addition to the spiral crimp) by passing through ~
rtuffer box, under m~ld gate ~nd roll pressure6. The resulting fiherfill has Q-crimp. The fiberballs of Examples 6-10 have 10-20% higher bulk than the ~iberballs of Exa~ples 1-5, whereas the molded product~ are not very different, but have lower resilience and lower Suppor~
~ulk (ss 60N).

-19- ~3~ Ei3~

~ xamples 11 and 13 6how the preparation of fiberballs with a preferred (non-silicone, hydrophilic) ~lickener being applied before ~he polyester fil~ments are relaxed, 60 as to "cure" ~he slicken~r onto the filaments during the relaxing treatment. The durability data of the resuiting cu~hions are compared in Table 3, w$th comparable products from the fiberfill of Example 1, while Tables 4 and 5 provide comparable data obtained from foam and latex products ~4) and from other molded fiber structures ~5) that are not accordinq to the invention.
ExamPle 11 A tow of ~symmetrically-jet-quenched drawn poly(ethylene terephthalate) filaments of 4.7 dtex was prepared conventionally without mechanical crimping, u~ing a draw ratio of 2.8X. The segmented copolymer sold as "~TLAS" G-7264, at a concentration of 0.35%, was applied to the fibers and dried at 130C. The tow was subsequently cut to 35 mm and relaxed at 1~5C. The staple was blended in the ratio of B0/20 with a sheath/core binder fiber, cut to the 6ame cut length, and having 4.4 dtex. The blend was opened on a commercial opener and the resulting opened fiber was processed into fiberballs e~sentially as described in Example 1.
The fiberballs were molded essentially as described in Example 1 into a cushion of 40~33X9 cm with a density of S0 Rg/m~. ~he cushion was ~ubmitted to the durability test described previously and the results 6how the improvement in durability versus Example 1, ~ainly with respect to the Work Recovery (resilience). The resilience losses of the product made according to Example 11 ~re about h~lf of the best Example in Table 2 with comparable bulk losses.
Example 12 This cushion was made with the 6taple of Example 1 for comparison with Example 11.

- r , -~3~63~

~xample 13 Thi6 was essentially like Example 11, except ehat the staple/binder r~tio was 90/10. Thi~ cu6hion 6hows excellent durability, but the resilience ~c much lower. Thi~ product has potential in back cu~hion~ or in styles requiring 60fter cushions.
Example 14 This cushion was ~ade with the ~taple of Example 1 hlended with the same binder at a ratio of 90/10, to ~ompare with Example 13. ~he durability te6t shows somewhat hiqher bulk 106ses than Example 12 tusing a ratio of 80/20).
Table 3 confirms that the re6illence of the molded 6tructures ~ade from the "dry" blend6 i~ higher than for the corresponding "61ickened" blends ~of ~xamples 11 ver~us 12, ~nd 13 versus 14). On the other hand, the molded fitructure~ made from the fiberball6 containing the "dry" blend have higher 106ses of resilience.
Comparison Products Table ~ ~hows the durability data for the following representative oam ~nd latex ~amples ~upplied by mattress ~nd furnishing manufacturers te~ted under the same conditions as the product~ of the ~nvention. Small differences between the initial v~lues of these products (as ~eported in Sable 4) ~nd the ~ea~urement6 reported previously (Table l) are ~ re~ult of 6ample to ~amp~e dif~erencefi or ~rom the ~ize of the 6ample. The re~ults in Table 4 are the measurement6 made on the piece actually tested cut to the came ~i~e as the molded cu6hion6:-Re 1: polyurethane foam of 30 Kg/m3 Re 2: polyurethane foam of 35 ~g~m3 n~oft"
Re 3: polyurethane foam of 35 Rg/m3 Re 4: polyurethane foam of ~0 ~g/m3 Re 5: latex matre~s core 72 ~g/m~
Table 5 6hows the comparable durability d~ta forcushions of the same ~ize from molded fiber ~tructures ; -20-~3~3~
that were not made from fiberball~, but ~lways used the s3me binder fiber.
Ct 1: 85/15 blend, u6ing 6 dtex hollow dry staple, carded, molded to a density of 50 ~g/m3.
Ct 2: Same blend, opened and random-filled into mold, same density.
Ct 3: Same blend, random-filled, but density of ~O Kg/m3.
Ct 4: Same as Ct 1, but th0 6 dtex hollow fiber had been coated with 0.35~ of the segmented copolymer of poly(ethylene terephthalate) and poly~ethylene oxide) a~
in Example 11.
Ct 5: Same blend as Ct 4, but opened and random-filled ~ike C~ 2.
Ct 5: Same as Ct 5, but den~ity only 40 Rg/m3.
The data contained in Tables 3, 4 and 5 c~n be ~nalyzed as follows:
Fiber assemblies made of blends of fiberfill binder in the appropri~te ra~ios, can produce molded cushions or ~imilar product~ with a durability which is better than foam and comparable to latex, at a comparable support bulk, by using fiberballs according to the invention.
The cushion, or mattress core, made from the fiberball~ of the invention has an import~nt advantage over foam and latex in having a higher air perme~bility than mo~t foam and latex, and a better moisture transport, due to the hydrophilic character of the "61ickener" and to the fi~erball structure.
Tne fiberball-molded cushions of the invention have 12-22% higher support bulk, but comparable to better durability ~t the same density, as compared with molded cushions made from conden6ed batts. Furthermore, a cushion molded from a earded batt does not adapt itself well to the human body. When a pre~sure is applied to its center, it pu116 thç ~ides, causing them ~o raise up. The ~ ~/D63~
~ushion made from the fiberballs of the invention adapt~
~tself to the deformation caused by the u~er, like a system comoposed of independent springs.
~ hese properties ~ake the products of the invention a much better product for furnishing cushlons, mattress cores, and 6imilar product6.
ProductG made from fiber blends such as th~ one used in Ct 4 have their own merits, particularly at lower densitie6 ~nd are the cubject of copending patent application DP-4155.
Individually-Bonded Fiberballs, e~. for ~illows In Examples 15-17, the fiberballs were not ~olded together to form an integral block, but were bonded individually, 60 that they can be u~ed as a highly-performing filling in r~fluffable cushions and pillows.
Bonding of the individual fiberballs can for instance be done in ~ fluidized bed.
~ n Examples 15 and 16, the fiberball6 of the invention were individually bonded, and then blown into a pillow ticking. In Example 17, for comparison, the fiberballs were not heated, i.e. were blown into the ticking without fir~t effecting bonding of the binder fiber~, In Ct. 1~, a commercially available beddin~
product (without binder fiber), the sub~ect of U.S. Patent No. 4,61~,631, was blown into the ticking to provide ~
further comparison. In each case, 10009 of the fiberball6 were ~illed into ~ ticking of dimensions 80 cm x 80 cm, and the compression measurements were made before and after flexing. Unlike the flexing used hereinbefore, however, the durability was tested using a Fatigue Tester as described in eolumns 9-10 of U.S. Patent No. 4,618,631, .
except that the severity of the flexing was increased to the extent that bulk losses after 6,000 cycles (in the present tests) csrrespond approximately to the bulk losses that had been sbtained after the full 10,000 cycle$ (as reported in U.S. ~atent No. 4,618,631) and the flexing was , ~L3C3~3~

continued (in the pre6ent te6t6) for a tot~l of 10,000 cycles; 60, it will be appreciated that the results reported in Table 6 reflect these ~ore ~evere flexing conditions than were used in u.s. Patent No. 4,618,631.
Example 15 The fiberball~ of the invention were produced as described in Example 1. The individual fiberball6 were then thinly-di6tributed bet~een two 6heets of a very open woven cotton fabric and heated in an oven at 160C. The fiberballs were thus essentially individually bonded ~any ball~ that were bonded together were ~eparated by hand).
1,0009 were thsn filled into the BOx80 cm pillow ticking by blowin~.
Example 16 The $iberball~ described in ~xample 15 were ~prayed with 0.35~ of the ~egmented copolymer 601d a~
"~TLAS" G-7264, dried at room temperature, and heated at 160C under the ~ame conditions as in Example 1~. The results in Table 6 show better retention of initial height than for the product of Example 15.
Example 17 The fiberballs were produced from the 6ame blend as in Example 15, but were not heated ~o the unbonded product was filled into pillow ticking and te~ted a6 a control for ExAmple 15, tc ~how the lmprovement achieved by bonding on the durability of the fiberb lls.
The data in Table 6 zhow:-The dry fiberball6 (Example 17) have a poorerdurability than the 61ickened commercial product (Ct 18), particularly at the ~upport bulk level ~60N).
"Dry" bonded fiberball~ (Example 15) ~how improved durability ver~us the 61ickened nonbonded commercial product (Ct 18), are much firmer and do not have the characteri6tic6 of a bedding product.
The fiberball~ which were slickened ~nd bonded IExample 16) 8how the best durability. ~he two bonded ;

`:
.

. .

2~-~30~3~

~amples (Examples 15 and 16) have much higher bl~lk than the nonbonded samples ~fter the durability test, which will translate into better-looking, more comf~rtable and ~ltogether more desirable furni6hing cu6hions.

.

~ ~ .

~ : ' ..
: ` !
, :

-25- ~30G34~

Table 1 Compression Characteristics Referred to ln Example 1 Sample ~ens~ty I82 B 7.5N SB 60N AS RS ~R
Kg/m cm cm cm cm ~ X
Invention A 20 9.0 7.25 3.5 1.75 19 64 B 25 8.45 7.45 4.35 1.0 12 71.5 C 35 8.15 7.3 5.0 0.85 10 74.5 : D 45 9.2 8.65 6.7 0.55 6 74 E 50 9.2 8.95 8.3 0.25 3 84 Comparisons 1 20 14.8 11.2 4.~ 3.6 24 63 2 25 12.4 9.45 3.6 2.95 24 63 3 35 9.7 9.6 7.2 0.1 1 75 4 30 9.6 9.5 8.0 0.1 1 76 72 10.5 10.1 8.0 0.4 4 81 ;

: -25-~ .

-26- ~3~

Table 2 Durability Data Com~ring ~xamples 1-10 I~2 B 7.5N SB 60N YPX
BF AF ~Z BF AF ~ BF AF dX BF AF ~

Ex. 1 9.2 8.95 2.7 B.95 8.65 3.4 8.3 7.85 5.483.8 71.45 14.7 Ex. 2 9.8 9.55 2.6 9.55 9.35 2.1 B.7 8.4 3.484.95 72.95 14.1 ~x. 3 9.3 9.2 1.1 9 B.8 2.2 7.55 7.25 4.0 66.25 60.45 8.8 ~x. 4 g.4 9.25 1.6 9.1 8.9 2.2 7.95 7.55 5.073.45 64.5 12.2 Ex. 5 9.2 8.95 2.7 9 8.75 2.B 8.4 8.1 3.682.3 75.0 8.9 ~x. 6 9.1 8.85 2.7 8.7 8.45 2.9 7.2 7.05 2.174.3 67.1 9.7 Ex. 7 9.35 9.05 3.2 9.05 8.85 2.2 8.05 7.9 1.9 B4.273.95 12.2 ~x. 8 9.15 8.95 2.2 8.85 8.65 2.3 7.95 7.6 4.4 82.070.25 14.3 ~x. 9 9.55 9.25 3.1 9 B.8 2.2 7.3 7.0 4.1 68.662.35 9.1 Ex. 10 9.1 8.85 2.78.85 8.6 2.8 7.95 7.65 3.8 79.7 72.8 8.7 ..
. .
,: .

.

~3C1 63~5~

Table 3 Durability Data of ~xamples 11-14 I~2 B 7.5N SB 60N ~RX
BF AF ~ BF AF ~X BF AF ~X BF AF ~X
Ex. 11 9.3 9.1 2.2 9.1 8.9 2.2 8.5 8.2 3.5 79.4 75.6 4.8 Ex. 12 9.1 8.8 3.3 8.8 8.5 3.4 8.1 7.7 4.g 84.6 74.1 12.4 Ex. 13 9.2 8.8 4.4 8.9 8.6 3.4 8.0 7.5 6.2 65.9 66.5 tl.O
Ex. 14 9.3 8.9 4.3 9.0 8.6 4.5 9.0 7.6 5.0 77.3 68.8 11.0 `'` ' :

' ., -28- 13Q63~

Table 4 Durabilitv Data of Reference SamDles (Foam and Latex) ~ . .
I~2 B 7.5N SB 60N ~RX
BF AF ~Z BF AF ~ BF AF ~X_ 8P AF ~X
Re 1 9.8 9.3-5.1 9.6 9.2-4.2 8.2 6.5 -20.7 79.3 71.8 -9.4 Re 2 9.9 9.6-3.0 9.7 9.5-2.1 7.3 5.8 -20.5 75.1 75.1 0 Re 3 10.2 9.8-4.010.0 9.8-2.0 8.7 7.6 -12.7 85.1 78.0 -8.3 Re 4 9.8 9.6-2.0 9.6 9.5-1.0 7.0 5.7 -18.6 85.8 86.7 -1.0 Re 5 10.3 10.2 -1.0 10.1 10.0 -1.0 B.1 7.8-3.7 80.1 80.1 0 ~' .~ .

'. ' ~ ' ~

, , .

-2g~ 63~

Table 5 Durability Data of Other Holded Fiber Cushions I~2 B 7.5N SB 60N YRX
sF AF QY BF AF ~X BF AF AY BF AF ~X
Ct 1 8.9 8.6 -3.4 8.6B.3 -3.5 7.57.2 -4.0 77.7 74.9 -3.6 Ct 2 9.0 8.6 -4.5 B.78.3 -4.6 7.46.9 -6.8 67.6 67 -0.9 Ct 3 9.0 ~.6 -6.7 8.67.9 -8.1 7.06.1 -12.6 65 64.1 -1.3 Ct 4 8.8 8.4 -4.5 8.38.0 -3.6 6,66.1 -7.6 76.6 76.3 -0.4 Ct 5 9.3 8.9 -4.3 8.57.9 -7.1 5.95.2 -11.9 72.7 71.2 -2.1 Ct 6 8.8 8.5 -3.4 B.17.4 -8.6 5.74.8 -15.8 74.2 72.3 -2.6 . :

:: :
- ` `'' ` ~ .
:

: .

~3~3~

Table 6 D bility Data of Pillows ~ ose Individually-Bonded) IH2 B 7.5N SB 60N URX
BF AF Q~ BF AF ~ BF AF ~X BF AF kX

Ex. 15 18.8 15.9 15.4 10.4 8.914.4 6.3 5.6 11.148.3 46.4 3.9 Ex. 16 17.7 15.6 11.9 9.B 8.5 13.3 6.3 5.6 11.146.9 44.4 5.3 Ex. 17 19.3 14.7 23.8 11.0 8.1 26.4 6.5 5.1 21.546.1 46.2 +0.2 Ct 18* 15.9 12.5 21.4 7.0 6.1 12.9 3.9 3.6 7.7 53.4 52.1 2.6 *Ct 18 was commercial fiberballs of slickened fiberfill, and of comparable density, but ~ithout binder.
+ The flexing was carried out appropriately for pillows, as descrlbed herein, snd not as described for firmer cushions.

` ~ -30-~`~
. ~
.... .
:

, ~

Claims (4)

1. Fiberballs of average dimension about 2 to about 15 mm consisting essentially of randomly-arranged, entangled, spirally-crimped polyester fiberfill having a cut length of about 10 to about 100 mm, intimately blended with binder fibers in amount about 5 to about 50%
by weight of the blend.

2. Fiberballs according to Claim 1, wherein the binder fibers comprise bicomponent fibers of cut length about 10 to about 100 mm, one component of which is binder material, whereas another component is polyester fiber of melting point higher than that of the binder fiber.

3. Fiberballs according to Claim 2, wherein the binder material comprises about 5 to about 50% of the weight of the bicomponent fiber.

4. Fiberballs according to Claim 2 wherein the binder fiber is spirally-crimped.5. Fiberballs according to Claim 3 wherein the binder fiber is spirally-crimped.6. Fiberballs of average dimension about 2 to 15 mm, consisting essentially of randomly-arranged, entangled, spirally-crimped bicomponent polyester/binder material fibers, having a cut length of about 10 to about 100 mm.
7. Fiberballs according to Claim 1, wherein the fiberfill has a coating cured thereto of a slickener consisting essentially of chains of poly(alkylene oxide).
8. Fiberballs according to Claim 1, wherein the fiberfill is coated with a segmented copolymer of poly(ethylene terephthalate) and poly(ethylene oxide) in amount about 0.05 to about 1% of the weight of the fiberflll.
9. Fiberballs according to Claim 1, wherein the fiberfill is coated with a modified poly(ethylene oxide)/poly(propylene oxide) granted with functional groups to permit crosslinking.
10. Process for making polyester fiberballs from a blend of spirally crimped polyester fiberfill with binder fibers, wherein small tufts of the blend are repeatedly tumbled by air against the wall of a vessel to provide the fiberballs.
11 Process for making polyester fiberballs from spirally-crimped bicomponent polyester/binder material fibers, wherein small tufts of the spirally-crimped fibers are repeatedly tumbled by air against the wall of a vessel to provide the fiberballs.
12. Process according to Claim 10, wherein the tufts are tumbled against a cylindrical wall of a vessel by air stirred by blades attached to a shaft rotating axially in the vessel.

13. Process according to Claim 11, wherein the tufts are tumbled against a cylindrical wall of a vessel by air stirred by blades attached to a shaft rotating axially in the vessel.
14. Process according to Claim 12, wherein the small tufts and the air are recirculated through the vessel.
15. Process according to Claim 13, wherein the small tufts and the air are recirculated through the vessel.
16. Process according to any one of Claims 10-15, wherein the tufts are formed by feeding loose fibers into the vessel, and by rotating the shaft and blades at a speed such that the fiberfill is separated into the small tufts.
17. Process according to any one of Claims 10-15, wherein small tufts that are not elongated are formed before feeding them into the vessel for rounding and condensing by air-tumbling.
18. Process according to any one of Claims 10-15, wherein the fibers are treated with a slickener to reduce the hairiness of the resulting fiberballs.
19. Process according to any one of Claims 10-15, wherein small tufts that are not elongated are formed before feeding them into the vessel for rounding and condensing by air-tumbling and the fibers are treated with a slickener to reduce the hairiness of the resulting fiberballs.
20. Process for making a bonded product, wherein an assembly of fiberballs according to any one of Claims 1 to 9 are heat-bonded and cooled.
21. Process for making a bonded product, wherein an assembly of fiberballs according to any one of Claims 1 to 9 are heat-bonded and cooled and wherein the fiberballs are first mixed with random binder fiber before forming an assembly and heat-bonding.
22. Process for making a bonded product, wherein an assembly of fiberballs according to any one of Claims 1 to 9 are heat-bonded and cooled and wherein the assembly is heat-bonded in a mold so as to produce a molded structure.
23. Process for making a bonded product, wherein an assembly of fiberballs according to any on of Claims 1 to 9 are heat-bonded and cooled and wherein the fiberballs are first mixed with random binder fiber before forming an assembly and heat-bonding and the assembly is heat-bonded in a mold so as to produce a molded structure.
24. Process for making bonded fiberballs, wherein individual fiberballs according to any one of Claims 1 to 9 are individually heat-bonded and cooled.
25. Process for making bonded fiberballs wherein individual fiberballs according to any one of claims 1 to 9 are individually heat-bonded in an airstream and cooled.
26. Process for making a loose bonded assembly, wherein bonded fiberballs are made according to any one of Claims 1 to 9, individually heat-bonded and cooled, and then assembled in a ticking.
27. Process for making a loose bonded assembly, wherein bonded fiberballs are made according to any one of Claims 1 to 9, individually heat-bonded in an airstream and cooled, and then assembled in a ticking.
28. A loose bonded assembly prepared by a process wherein individual fiberballs according to any one of Claims 1 to 9 have been individually heat-bonded and cooled.
29. A loose bonded assembly prepared by a process wherein individual fiberballs according to any one of Claims 1 to 9 have been individually heat-bonded in an airstream and cooled.
30. Individually-bonded fiberballs prepared by a process wherein individual fiberballs according to any one of Claims 1 to 9 have been individually heat-bonded and cooled.
31. Individually-bonded fiberballs prepared by a process wherein individual fiberballs according to any one of Claims 1 to 9 have been individually heat-bonded in an airstream and cooled.
32. Molded structures prepared by a process for making a bonded product wherein an assembly of fiberballs according to any one of Claims 1 to 9 are heat-bonded and cooled and wherein the fiberballs are first mixed with random binder fiber before forming an assembly and heat-bonding and the assembly is heat-bonded in a mold so as to produce a molded structure.
33. Molded structures prepared by a process for making a bonded product wherein an assembly of fiberballs according to any one of Claims 1 to 9 are heat-bonded and cooled and wherein the fiberballs are first mixed with random binder fiber before forming an assembly and heat-bonding and the assembly is heat-bonded in a mold so as to produce, said structures having a density of from about 20 to about 80 kg/m3.